This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Television host Chris Cuomo shared an AI-generated deepfake video of congresswoman Alexandria Ocasio-Cortez on Instagram, mistakenly treating it as a genuine clip to criticize her. Despite the video being clearly watermarked as a deepfake, Cuomo used it to fuel political commentary. After being corrected by Ocasio-Cortez, Cuomo deleted the post but continued to criticize her, leading to public mockery and a broader discussion on media literacy and the impact of AI-generated content in political discourse.
NewsNation host Chris Cuomo reportedly shared on Instagram a video of Alexandria Ocasio-Cortez that was prominently labeled as a parody deepfake created with AI technology. In his commentary, Cuomo purportedly treated the statements in the video as genuine, using them to criticize Ocasio-Cortez for his audience.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Users anthropomorphizing, trusting, or relying on AI systems, leading to emotional or material dependence and inappropriate relationships with or expectations of AI systems. Trust can be exploited by malicious actors (e.g., to harvest personal information or enable manipulation), or result in harm from inappropriate use of AI in critical situations (e.g., medical emergency). Overreliance on AI systems can compromise autonomy and weaken social ties.
Also classified as
The AI-generated deepfake video, despite being a parody, functioned as misleading information that deceived a public figure and spread to his followers.
The incident represents unsafe use and inappropriate reliance on unverified AI-generated content by a media professional, who failed to notice or heed the watermark indicating the video was a deepfake.
Human
Due to a decision or action made by humans
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was primarily caused by a human decision (Chris Cuomo sharing the video) and his unintentional blunder of mistaking a clearly marked parody deepfake for real news.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
A US television host mistakenly shared a watermarked deepfake of a congresswoman as real, highlighting the risks of AI-generated misinformation in political discourse. While the incident caused public embarrassment and minor epistemic confusion, it posed no significant threat to national security or critical infrastructure.
The AI system was used as a tool to generate a deepfake video based on direct inputs, performing a specific function of video synthesis.
Risk Level 3: Limited Risk. The incident involves an AI-generated deepfake video, which falls under the category of AI systems that require transparency obligations to ensure users are aware they are interacting with AI-generated content.
The AI system was used to generate a deepfake video of Alexandria Ocasio-Cortez, including synthetic voice generation that mimicked her speech.
The report clearly details the incident where Chris Cuomo shared a watermarked deepfake of Alexandria Ocasio-Cortez, her response, and his subsequent reactions. There is little ambiguity about the sequence of events.
None
Technology
Realistic Deepfake Generation
AI tools allowed creation of highly realistic parody video of the politician.
Watermark Obscurity
Though watermarked, the AI indicator did not prevent user sharing or belief.
Human Factors
Lack of Critical Thinking
The host shared the video without verifying its authenticity or watermark.
Confirmation Bias
Host readily believed the fake video because it aligned with political views.
Failure to Notice Watermark
The user ignored the prominent 'deepfake' watermark on the video.
Process and Methods
Inadequate Content Verification
Journalist failed to verify source before posting on social media.
Lack of Editorial Oversight
Personal social media accounts of journalists lack standard newsroom checks.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
